© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 92037
    Prev
    Next
    INDEX
    Explanations

    commas

    np_max-act · gemini-2.0-flash

    The neuron activates on numeric values and quantitative/descriptive adjectives (e.g. ages, sizes, “small,” “large,” “average,” “strong,” “inexperienced”), i.e. tokens that express measurable or physical attributes.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Eck
    -0.07
    03
    -0.07
     менше
    -0.07
     nested
    -0.07
    .ev
    -0.07
     shepherd
    -0.07
    ieu
    -0.07
    _TEM
    -0.06
     execut
    -0.06
    .IntPtr
    -0.06
    POSITIVE LOGITS
    �
    0.06
     charismatic
    0.06
    	elseif
    0.06
     Cumhurbaş
    0.06
     �
    0.06
    )*(
    0.06
    .QRect
    0.06
     FORMAT
    0.06
    itical
    0.06
    转
    0.06
    Activations Density 0.004%

    No Known Activations